Contributory inequality alters assessment of academic output gap between comparable countries

Abstract

a b s t r a c tAn elite segment of the academic output gap between Denmark and Norway was examined using harmonic estimates of publication credit for contributions to Science and Nature in 2012 and 2013. Denmark still leads but the gap narrowed in 2013 as Norway's credit increased 58%, while Denmark's credit increased only 5.4%, even though Norway had 36% fewer, and Denmark 40% more, coauthor contributions than in 2012. Concurrently, the credit produced by the least productive half of the contributions rose tenfold from 0.9% to 10.1% for Norway, but dropped from 7.2% to 5.7% for Denmark. Overall, contributory inequality as measured by the Gini coefficient, fell from 0.78 to 0.51 for Norway, but rose from 0.63 to 0.68 for Denmark. Neither gap narrowing nor the positive association between reduced contributory inequality and increased credit were detected by conventional metrics. Conventional metrics are confounded by equalizing bias (EqB) which favours small contributors at the expense of large contributors, and which carries an element of reverse meritocracy and systemic injustice into bibliometric performance assessment. EqB was corrected by using all relevant byline information from every coauthored publication in the source data. This approach demonstrates the feasibility of using EqB-corrected publication credit in gap assessment at the national level.

A review of the literature on citation impact indicators

“…These include harmonic counting (Hagen, 2008(Hagen, , 2010(Hagen, , 2013(Hagen, , 2014a(Hagen, , 2014b(Hagen, , 2015Hodge & Greenberg, 1981;Jian & Xiaoli, 2013), arithmetic counting (Abbas, 2011;Egghe et al, 2000;Van Hooydonk, 1997), also known as proportional counting, geometric counting (Egghe et al, 2000), the counting method of Assimakis and Adam (2010) based on the golden number, and the axiomatic counting method of Stallings et al (2013). Table 4 illustrates the differences between these approaches by showing the weights assigned to the authors of a publication with five authors.…”